Low Liquidity Pairs : Is Trading Them Worth the Hassle?

I have always been interested in the analysis of different forex currency pairs and the difficulty or easiness in which we can find inefficiencies amongst them. Through years of studying the market I have encountered a very particular and quite obvious phenomena which relates strongly with pair liquidity. When building systems for different instruments in Forex trading it becomes clear that pairs which have higher liquidity are easier to profit from while pairs which have lower liquidity have inefficiencies which are more difficult to find and more prone to “going away” quickly. On today’s post I am going to talk about low liquidity pairs and why I believe that trading them might now be worth the over exposure and hassle involving low liquidity levels.

It is true that we cannot know the real volume of any given instrument on the current over-the-counter spot Forex market. However by looking at the records of brokers with “open books” it becomes quite clear that the large majority of money in the Forex market flows through 4 to 6 different currency pairs. This means that most pairs available have actually low volumes of trading, sometimes being even ridiculous when compared with the massively traded pairs. For example the traded volume of a pair like the GBP/CAD can be almost a thousand times smaller than the volume traded on a pair like the EUR/USD (which by some estimates can account for about half of all the Forex trading volume).

The effect of more volume – more liquidity – is simply a more technically well-behaved pair. When volume is low currency pairs start to show the same problems as low liquidity stocks, their behavior can reflect the trading of a few individuals and therefore they are not only more prone to market manipulation but it becomes difficult to see true inefficiencies as there is a lot of “noise” which we cannot get through. Market inefficiencies seem to arise from the manifestation of the true nature of collective behavior and such manifestation is low or almost non-existent on pairs which have lower liquidity.

Another problem of low liquidity pairs is definitely the lack of data available for them. For most minors intra-day data only goes back a few years limiting the development of systems to those that rely solely on daily data which can go back even to the early 1990s. We also need to take into account that the little intra-day data we have isn’t usable since low liquidity pairs tend to have large historical spread variations which require us to have adequate Bid/Ask tick feeds to know when and how this happened.

Certainly it is hard to say that it is “easier” to develop systems on one pair or another since this would – in essence – relate only with the experience of the person making this statement (in this case me). However I have tried to detach myself from a personal opinion or skill level by using genetic programming frameworks to develop systems on currencies with both high and low liquidity levels. Not surprisingly, frameworks like Coatl are able to find profitable systems for major currency pairs quite easily (with very good AMR to Max DD ratios) while minors are more troublesome and require more time and effort to find even only mediocre systems which aren’t – most of the time – even half as profitable as systems developed on more liquid currency pairs.

The problem here is therefore completely related with pair liquidity since the same level of data available for several different pairs (only daily data) can yield very profitable systems on a highly liquid pair and very low profitability systems on the low liquidity part of the spectrum. Therefore the question arises of whether or not it is worth it to explore and create systems outside the 5 or 6 most liquid forex currency pairs. The answer – from a profit stand point – seems to be that it makes no sense explore less liquid pairs since their results are much less profitable than for liquid instruments.

Another problem comes when you do some stress tests to see how successful a system might be under out-of-sample testing conditions. When building systems using a period of – for example – 5 years, highly liquid pairs tend to show much more success within the next five years while low liquidity pairs tend to fail a big percentage of the time. This means that the underlying inefficiencies found on low liquidity pairs are often the result of some behavior which is not inherent to them while on highly liquid pairs crowd behavior is always present and therefore inefficiencies tend to last for longer periods of time as they are “seen more clearly” than on minors.

Overall experimentation with genetics and out-of-sample testing has revealed that trading systems developed on minors are often less profitable and more prone to failure than systems developed for highly liquid currency pairs. However it is true that portfolios built over minors might offer some “relief” from these problems and in the long term – as the liquidity of these pairs increases – the profitability of inefficiencies found within them might start to greatly improve. As I have said before I have the utmost intention of building genetic engineered portfolios for minors, something which will put the above assumptions to the harshest test — out of sample live testing.

If you would like to learn more about my work in automated trading and how you too can learn to develop systems for highly liquid and low liquidity pairs please consider joining Asirikuy.com, a website filled with educational videos, trading systems, development and a sound, honest and transparent approach towards automated trading in general . I hope you enjoyed this article ! :o)

Print Friendly, PDF & Email
You can leave a response, or trackback from your own site.

Leave a Reply

WordPress › Error

The site is experiencing technical difficulties.